Convergencestructure after the object

Instrument lineage: the ISR library: Matisse, Degas, Pollock, Caravaggio, Soejima, Klimt, Bruegel, Cartier-Bresson, Sora Pair, MidJourney Ensemble, Velazquez, and Malevich

Context

A late drip painting under pressure

Convergence belongs to Pollock's mature poured-paint period, after the breakthrough all-over canvases of the late 1940s and after the more restricted black-and-white paintings that followed. The common account matters here because it gives the painting a built-in tension: it begins from a black-and-white armature, then receives red, yellow, blue, and white interventions that make the field more chromatic without giving it a conventional figure.

The usual formal language around the work is familiar: shallow optical space, line released from contour, compressed layering, all-over composition, and a palette held together by black and white. Our analysis keeps those claims in view, but tests where they become too broad. Earlier composition concentrated consequence into privileged locations: horizon, figure, focal point, object boundary. Convergence removes those anchors but does not remove organization. It redistributes consequence.

Claim 1

All-over does not mean undifferentiated

The first correction is to the phrase "all-over." The structure reads as a TEXTURE_FIELD, with low cohesion μ=0.166 and very low orientation stability θ=0.0032. That supports an all-over field in the structural sense. But role-by-border testing shows that the field is not equal in every channel. Edge density is border-active, black is border-active, and canvas/void is border-active. Strict primary chroma retreats.

The frame profile makes the split measurable. Pollock's outer edge density is high relative to the inner field, but primary-color density is lower near the edge: primary outer/inner ratio 0.805, primary pullback 0.195. So the painting is all-over structurally, not all-over chromatically.

Claim 2

The border is a behavioral problem, not just an edge problem

The protected-border observation becomes more interesting when phrase families are tested separately. Eligible phrase events have a perimeter-band share of 0.359 and an outer-8% border share of 0.196. Against that baseline, blue/white behaves as protected_or_refusing, red/orange behaves as border_tolerant, and black/white/yellow behaves as baseline_like.

This gives the critique a stronger sentence: Pollock does not simply protect the edge. He assigns different edge permissions to different kinds of local event. Blue/white works as an interior release system. Red/orange is the more border-tolerant disruptor. Black/white/yellow is the baseline stage that can operate broadly across the field.

Claim 3

Color advances, but not by escaping contour

The push/pull hypothesis survives, but only in a harder form. Chroma is perceptually active: strict primary chroma saliency lift is 1.41× and broad chroma saliency lift is 1.37×. But the strict chroma Canny capture share is 0.814, so color is not cleanly soft. It is contour-entangled.

The better reading is that Pollock makes push/pull out of interlock. Saturation pulls color forward. Black stabilizes the field as scaffold and table. White interrupts black and lifts passages toward the eye. Canvas/void remains active as spacing and depth medium. The painting does not separate hard drawing from soft color. It forces them to trade roles.

Claim 4

The handoffs are the spatial engine

The transition-zone results give us a way to talk about depth without pretending to know the physical paint sequence from the reproduction. The transition union is broad, but it is not mainly a hard contour field. Its Canny share is 0.258, and the softest interface is broad_chroma_to_canvas. That means the spatial action lives in handoff zones more than in isolated color territories.

This matters for the figure-ground question. Color can become figure, but it keeps sliding into canvas, black, and white. Canvas is not passive background. Black is not pure depth. White is not just highlight. The roles change at their borders, and those borders are frequent enough to become the painting's engine.

Claim 5

The colors behave more like rhythm than direction

The directional color probe weakens the idea that each color carries a stable arrow. All orientation coherence values are low. Red/orange is the most directional role at 0.036, but even that is not strong enough to justify a simple directional reading. The rhythm probe gives a better language: roles recur as pulses, spacing intervals, and local contacts.

The stress test supports recurrence rather than random scatter. Yellow's mid-scale Moran z is 3.19, black's is 3.87, and white's is 2.99. Direct contacts are often depleted against shifted-role controls, which is important. Pollock's roles do not just collide. They pulse near one another, miss one another, and keep intervals open.

Claim 6

All-over organization becomes local recurrence

The later probes move from roles to recurring local behavior. The interval network is connected as a union, but individual pair corridors fragment. Junction testing then asks what happens at the switches. The phrase probe reduces those switches to local clauses for measurement only: corridor arm A, junction signature, corridor arm B. The result is phrase_grammar_supported, with 1449 eligible junctions and 2669 arm-to-arm clauses.

The dominant recurring clause is black_to_white__red_orange_to_yellow|black+white+yellow. That does not make the painting readable as a language. It shows something earlier and safer: the field has measurable recurring structural behaviors without objects. The old language of "liberated line" can now be tightened. The line is not only free from contour. It repeatedly enters similar local exchanges.

Claim 7

The viewer gets stations, not a single path

The saliency and route probes keep the critique honest. They do not prove eye tracking. They test whether repeated phrase structures are likely to become attention events. Phrase saliency coupling is phrase_saliency_coupling_supported. The route graph is mixed: active nodes form local circuits, but the largest component share is only 0.08. That argues against one continuous eye-path highway.

The station pass gives the positive version of that failure. It finds 44 active route components, 23 station-sized components, and station node share 0.72. The viewer is not forced along one route. The viewer is caught, switched, released, and caught again.

What Can Be Said

A stronger but narrower critique

The evidence supports a critique of Convergence as a distributed system rather than a homogeneous field. It is all-over in structural reach, not in equal chromatic behavior. It protects chroma at the edge, but not every phrase family equally. It creates push/pull, but not by keeping color soft and line hard. It creates rhythm, but mostly through recurrence, interval, and near-miss spacing rather than through clean direction.

Public Frame

Where this sits against existing readings

This critique keeps several public art-historical frames in the background. Museum writing on Pollock's 1950 poured paintings emphasizes the floor-laid canvas, movement around all sides, poured or flung paint, dense interlacing, and the paradox of apparent freedom with underlying order. The Met's account of Autumn Rhythm also stresses Pollock's relation to mural scale and the speed with which his reinvention of painting entered the modern canon. Those readings support the broad setup here: Pollock's drip field is process-driven, wall-scale, and anti-easel in ambition.

The critique also sits between Greenberg and Rosenberg without adopting either whole. Greenberg's formalist problem is useful for all-overness, optical field, and the crisis of easel composition. Rosenberg's action-painting problem is useful for process, event, and bodily performance. Our evidence does not settle that historical debate. It suggests a bridge: the action leaves behind measurable recurring structural behavior.

Finally, the fractal literature is treated as context, not as proof. Pollock's work has often invited fractal and multifractal analysis, but later critiques warn that fractal measurements can be fragile and should not be used as simple authentication or intention claims. That is why this critique uses self-similarity as one supporting behavior among many, not as the master explanation.

Sources consulted: MoMA, Jackson Pollock, One: Number 31, 1950; The Metropolitan Museum of Art, Autumn Rhythm: Number 30, 1950; summary catalog data for Convergence; Harold Rosenberg, "The American Action Painters" (1952); Clement Greenberg's writings on all-over painting and American-type painting; Yve-Alain Bois, "Whose Formalism?" (1996); Katherine Jones-Smith, Harsh Mathur, and Lawrence M. Krauss on limits of fractal authentication; Roberto Zenit on fluid mechanics and Pollock's controlled viscous filaments.

What This Does Not Prove

Limits of the evidence

This is a reproduction-based analysis. It cannot prove Pollock's intention, physical drying sequence, exact pigment behavior, conservation history, or the viewer's actual eye movements. The masks are evidence tools, not the painting itself. The strongest claims are therefore behavioral: this image, at this resolution, shows structural all-overness with chromatic pullback, role-specific border behavior, contour-entangled chroma, transition-heavy depth, recurring local clauses, and station-like attention pressure.

n = 1, and one reproduction at that. Every number is from a single image. JPEG compression, the photographer's lighting, and color profile all directly affect edge detection, chroma thresholds, saliency, and box-counting. The analysis knows this: it keeps raw vs edge-aware checks where useful and flags the border color read against reproduction artifacts. But no single-image analysis can fully separate a fact about the painting from a fact about this photograph of it.